I guess I’m imagining that in a world where we only “barely tolerate” the pain, many people still try to defect over the long run due to psychological buildup, and end up getting killed. Thus there is some pressure to do better than “barely tolerate” at least. I’ve definitely updated to a model where the shock still produces suffering though; I no longer think the evolutionary pressure would be sufficient to remove the pain entirely.
The key is that evolution would try to find an equilibrium between the reproductive risk of not shocking yourself enough, and the reproductive risk of shocking yourself too much. Since not shocking yourself has much higher consequences on the margin, I expect evolution to bias slightly towards shocking yourself too much.
I guess I’m imagining that in a world where we only “barely tolerate” the pain, many people still try to defect over the long run due to psychological buildup, and end up getting killed. Thus there is some pressure to do better than “barely tolerate” at least. I’ve definitely updated to a model where the shock still produces suffering though; I no longer think the evolutionary pressure would be sufficient to remove the pain entirely.
The key is that evolution would try to find an equilibrium between the reproductive risk of not shocking yourself enough, and the reproductive risk of shocking yourself too much. Since not shocking yourself has much higher consequences on the margin, I expect evolution to bias slightly towards shocking yourself too much.